Method for increasing resolutions of depth images

Abstract

A resolution of a low resolution depth image is increased by applying joint geodesic upsampling to a high resolution image to obtain a geodesic distance map. Depths in the low resolution depth image are interpolated using the geodesic distance map to obtain a high resolution depth image. The high resolution image can be a gray scale or color image, or a binary boundary map. The low resolution depth image can be acquired by any type of depth sensor.

Description

FIELD OF THE INVENTION[0001]The invention relates generally to depth images, and more particularly to increasing resolutions of depth images.BACKGROUND OF THE INVENTION[0002]Depth images can be used for 3D scene analysis and other 3D computer vision applications. Typically, the resolution of the depth images is substantially less than comparable optical images, i.e., gray scale or color images. The resolution of images acquired by optical cameras can easily be ten megapixels or greater, while the resolution of depth images is typically less than 0.02 megapixels. As defined herein, pixels in optical images only have associated intensities, perhaps in multiple color channels. A depth image only has associated depths for the pixels.[0003]One way to increase the resolution of the depth images is to use a high resolution optical image that is registered to a low resolution depth image. In general depth and color boundaries of a scene are correlated. Abrupt depth transition often leads to...

AI Technical Summary

Benefits of technology

[0009]The method uses a novel approximation procedure with a linear complexity as a function of image size and can achieve real-time performance. In addition, the method can also be used with binary edge maps in sensor fusion applications.

Problems solved by technology

Hence, obtaining geodesic distances by solving the all-pair-shortest-path problem is unsuitable for real-time processing.

Images